|dc.description.abstract||Dielectric oils are frequently used as an insulating medium in high-voltage components such as transformers, capacitors and cables. Improving their insulating systems are essential to improving their performance. Dielectric behavior between such components are guaranteed by using a liquid or solid-liquid interface which is designed to quickly suppress electrical discharges.
Recent experiments have detected high amplitude photomultiplier pulses from 2nd mode streamers. However, it remains unclear whether they are caused by the spectral characteristics of the photomultiplier or if they are a consequence of stepped propagation of the streamer. In the experiments performed in this work, the current and emitted light of 2nd mode streamers in cyclohexane with pyrene and butylated hydroxytoluene as additives were studied. Photomultipliers were used to measure the emitted light intensity and a new experimental setup was designed to measure the current. The test-cell had a point-plane geometry with a gap distance of 4-6mm, needle radius was 15μm and an impulse generator was used to generate rectangular high-voltage pulses with typical risetimes of approx. 20ns.
Investigations of the photomultiplier outputs found that the pulse amplitudes are much higher than the dark-current, which corresponds to a single electron pulse. This was taken to make it less probable that continuous light is detected as pulses due to a combination of weak light and the quantum efficiency of the photomultiplier. Furthermore, a qualitative comparison of two different signals from the same streamer found that some pulses appear at the same time in both signals.
Butylated hydroxytoluene was found to have similar effect on a liquid s dielectric properties as pyrene. Breakdown voltage is reduced, and acceleration voltage is increased, making the 2nd mode appear over a larger voltage range.
A setup designed to measure pulses in the current through the test-cell has been proposed. The measurements found no correspondence between current and photomultiplier pulses; however, some sources of errors have been discussed which can be used to further improve the setup for further investigations.||